These days people increasingly treasure leisure activities. As a result recreational businesses are flourishing. Among them gambling is very popular. Throwing or rolling dice is one of the widely loved games. It is almost ubiquitous, and can be seen in houses, outdoors and business environments. One of the common rules of the game of rolling dice is: players place bets first and guess the total dots; the dealer rolls the dice and counts the total dots when moving of the dice stops to determine win and loss.
In the conventional game playing rolling of the dice is done manually. The rectitude or fairness of the dealer who rolls the dice often is suspected by the players. As the force or angle of throwing the dice could affect the dot number and outcome. Some even resort to cheating schemes such as crooked dice or loaded dice to take the advantages of the ignorance of the players. This spoils the pleasure of playing dice games.
To remedy the aforesaid problem, automatic dice rolling machines have been developed that have a recognition means to confirm dice dots. The recognition means usually performs recognition by detecting the dots through electric signals such as picture signals. However, such a method is easily compromised by control of people and produces fake results. This creates concerns of the players and affects their willingness and interests in playing the game. Hence the automatic dice rolling machine is not well accepted in legitimate recreational sites.
Moreover, even if the dealer does not tamper the dice and affect the fairness of the dice games, the automatic dice rolling machine is prone to recognition errors due to deficiency of recognition techniques, such as starting recognition process while the dice are still moving, impact of environmental factors such as light and angles of image capturing and the like.
Therefore how to provide an accurate method to recognize dice dots to improve the efficiency and fairness of dice rolling game is an issue remained to be resolved in the industry.